Transcriptome and Mendelian randomization were combined to screen and validate prognostic genes associated with lipid autophagy in oral squamous cell carcinoma

Background Cancer cells can use fatty acids produced by lipophagy to continue growing and proliferating, but the regulation of lipophagy in oral squamous cell carcinoma (OSCC) remain poorly understood. Methods mRNA profiles, expression quantitative trait loci (eQTL) data, and ieu-b-4961 were scoured from the online databases. In TCGA-OSCC, the differentially expressed genes (DEGs) were screened between the tumors and paracancerous tissues. The weighted gene co-expression network analysis (WGCNA) was applied to obtain the key module genes highly related to lipophagy. Later, differentially expressed lipophagy-related genes (DE-LRGs) were determined by overlapping DEGs and key module genes. Next, the eQTL data of DE-LRGs was an exposure factor and the OSCC was an outcome factor in the two-sample Mendelian Randomization (MR). Meanwhile, sensitivity analyses and MR Steiger filtering were performed, and then candidate genes were selected to construct a prognostic risk model. Based on least absolute shrinkage and selection operator (LASSO)-Cox regression analyses, the prognostic genes were confirmed and a prognostic risk model was built. Afterwards, the tumors of OSCC patients were divided into high- and low-risk teams based on the median risk score. Finally, the immune microenvironment was evaluated using the estimate and single sample gene set enrichment analysis (ssGSEA) algorithms. Results A total of 271 DE-LRGs were determined by overlapping 4,712 DEGs and 308 key module genes. Among them, 18 exposure factors could affect directly OSCC as candidate genes. Next, 4 prognostic genes ( CLTCL1 , TNNC1 , ALPK3 , and PFKM ) were identified, among them, CLTCL1 (odds ratio (OR) = 0.9997, 95% confidence intervals (CI) = 0.9995–0.9999, P _IVW = 0.0020), PFKM (OR = 0.9997, 95% CI = 0.9995–0.9999, P _IVW = 0.0067), and ALPK3 (OR = 0.9990, 95% CI = 0.9983–0.9997, P _IVW = 0.0061) were protective factors and TNNC1 (OR = 1.0005, 95% CI = 1.0001–1.0008, P _IVW = 0.0102) was a risk factor. A prognostic risk model was built, notably, the probability of overall survival (OS) in the low-risk team was higher than that in the high-risk team. Furthermore, we found that the low-risk team had higher immune, stromal, and ESTIMATE scores, and there were 23 differential immune cells between the two risk teams. Conclusion Generally, CLTCL1 , PFKM , and ALPK3 were protective factors, while TNNC1 was a risk factor for OSCC patients. Our findings provided a new perspective on the treatment and prognosis of OSCC.